Data Visualization Case (Laetitia H.)¶

image.png

If I were to answer the case with one graph...¶

here is how the recruitment process looks like

The different possible application stages are ... (alphabetical order)¶

  • Use case : as a recruiter, I want to select "Recruiter ap..." for an entry, which one should I select?
  • Use case 2 : as a recruiter, I just received an application, and I screened it, what stage should I update? Application Review? Recruiter Screening? Screened? Application Stage, even?
  • What about the 'delete' state?
  • If "Hired" can be a Stage, what is the stage to mark the Rejection?

Let's preprocess the data and take some bold hypothesis¶

  • Recruiter application (spelling correct)
  • Interviews (1,2,3) can be either F2F or Phone/Video -> group them
  • Test SHL is also Testtake home
  • //Recruiter application is also Application stage
  • Application Review = Hiring Manager Review
  • Delete 'Interview 1 - F2F - delete' stage
  • inbalanced created by delete stage_name and isDelete column removed
  • wait time until application ackowledged : df['entry_log_wait']
  • length per application unique number : df['length_per_application']
  • length per application_id and refined per status : df['length_per_application_stage']
  • sum of different application_id per candidate_id : df['sum_applications_per_candidate']
  • total number of stage_name per application_id & candidate_id : df['unique_stages_per_application_candidate']
  • flag the final process stage per application_id per candidate_id : df['exit_flag']

In CONCLUSION, the stage name is the key issue as it seems to provide confusion.
First of all, could clarify the process, its stages and what they mean ?
Yet, the process does also loop over and over.
We could analyze deeper to provide a more complete answer.

High level summary of the recruitment per status : rejected - active - hired¶

Agenda - Elements from the highlight summary to study :

  1. Number of applications by candidate
  2. Stages analysis per application
  3. Special case : 10 stages for this candidate
  4. Length of application - length of stages
  5. Department busy-ness evolution

1. Number of applications by candidate¶

2. Stages analysis per application¶

3. Special case : 10 stages for this candidate (extreme case)¶

1 candidate went through 10 stages of candidacy to be hired after 65 days since first entry log. That was his second application.

candidate_id job_id applied_at length_per_application entry_log_wait sum_applications_per_candidate
69 0x58E38E6EE17BF0549F3B3E2C1D20FEBC 1249559 2020-01-15 00:26:00 65.0 65 2

4. Number of application - length of stages¶

5. Department busy-ness evolution¶

In full conclusion ...
Is it the process that could be improved?
Or the systems to handle business evolution that needed to be scaled up?